Project Summary: Expanding the Scope of Base Editing Genome editing has revolutionized the life sciences and offers the potential to cure genetic diseases. We recently developed base editing, a method of making single-base changes at target genomic sites without introducing double-strand breaks or relying on homologous recombination. Base editors (BEs) are especially relevant for the study and treatment of genetic diseases because the majority of disease-relevant mutations are single-base changes. In the two years since we pioneered base editing with a C?G-to-T?A editor, we have improved BE efficiency and product purity, reduced off-target and bystander base editing, evolved a new class of adenine base editors (ABEs) that convert A?T to G?C base pairs, expanded the targeting scope of BEs, established base editing of post-mitotic somatic cells in vivo, and applied BEs to record cellular events. Hundreds of other laboratories around the world have used base editing to study genetic diseases and to test potential therapeutic strategies. Here we propose to expand the capabilities of base editors towards the transformative goal of enabling any desired base change at any target locus in any somatic cell. Base editing requires the presence of an appropriately positioned protospacer adjacent motif (PAM) for binding of the Cas9 domain. Most DNA sites remain inaccessible for genome editing due to the lack of any DNA-binding CRISPR protein that recognizes the majority of PAMs. To further expand our ability to base edit the broadest range of targets, we will use our phage-assisted continuous evolution (PACE) platform to rapidly evolve a collection of Cas9 variants that recognize currently many untargetable PAM sequences (Aim 1a). The targeting scope of BEs is also limited by inefficient editing of certain base pairs because of sequence context. To further expand the targeting scope of base editing, we will use our recently established PACE selection for base editing to generate BEs that can efficiently modify targets with currently disfavored flanking sequences (Aim 1b). Base editors modify bases within the editing window, a range of ~5 nucleotides positioned relative to the PAM. In addition to conversion of the target C?G or A?T base pair, other ?bystander? C?G or A?T base pairs are also edited within this window. These bystander edits can lead to undesired genome changes. To minimize bystander base editing, we propose to evolve a large set of BEs that will only edit bases within specific sequence contexts (Aim 2), thereby enabling discrimination between multiple Cs or As within the editing window. Finally, a major limitation of base editing is the inability to generate transversion (purine pyrimidine) mutations, which are needed to install or correct ~38% of known human pathogenic SNPs. We propose to develop the first base editors that can generate transversion mutations at target base pairs using two distinct strategies (Aims 3a and 3b). Success with either strategy would greatly expand the capabilities of base editing, and would also allow, in principle, all 12 possible base-to-base change via individual or sequential use of transition and transversion editors.